Research On Scenario-based Expression And Optimal Emergency Response Action For Marine Oil Spill Accident

Major marine oil spills are mainly due to offshore oil production blowout accidents,subsea oil pipeline damage,and marine ship collision.With economic growth,China’s dependence on oil is becoming stronger,and China has become the world’s secondlargest oil consumer.With oil imports exceeding 420 million tons in 2017,China has surpassed the United States for the first time and become the world’s largest oil importer.Based on the analysis of the occurrence of marine oil spill emergencies and unconventional emergencies,this article focuses on the use of historical cases for supporting decision-making in marine oil spill emergencies.Combining artificial intelligence,scenario analysis,and scenario safety entropy,this paper studies the expression of marine oil spill accidents based on the scenario,the measurement method of marine oil spills,and the optimal response strategy training for marine oil spill emergencies based on deep reinforcement learning(DRL).The key concepts and results of this work are as follows:1)After an in-depth analysis of marine oil spill emergencies,this research pointed out that marine oil spill emergencies are difficult to predict,have complicated evolutions and causes,produce wide-reaching environmental harm,and are difficult to remediate.Aiming at the above-mentioned characteristics,a scenario-based method of marine oil spill accident is proposed and is used to further study this topic.2)Focusing on the significant value of historical cases of marine oil spill accident in supporting decision-making,a semantic analysis-based framework to extract scenario element instances related to marine spill accident from text is proposed,which includes the conversion of “event” to “scenario”.3)Aiming at the existing support of decision-making systems for marine oil spills,which is insufficiently utilized in historical cases,this paper proposes an optimal strategy training method based on reinforcement learning.When applying the proposed method to train the model,it not only makes full use of the existing historical cases but also trains the purposeful coping strategy model in combination with the decision intention.4)Based on the above research results,an emergency decision-making system for marine oil spill emergencies was designed.On the premise of the distributed cluster architecture,the core module of the support system consists of the marine oil spill emergency case management module,the construction of marine oil spill emergency scenarios module,and the optimal response module.The above “Research on Scenario-based Expression and Optimal Emergency Response Action for Marine Oil Spill Accident” may make a useful exploration of the emergency management theory and emergency response technology of the “ScenarioResponse” paradigm,which is beneficial to the emergency response of marine oil spill accidents and the improvement in decision-making.